A control plane for downtime incidents, failure modes, and restart-safe recovery.

This repo models the operating layer between incident intake and restart execution: line visibility, root-cause mapping, evidence routing, restart blockers, and operator-safe handoffs.